For utilities, maximizing revenues in a competitive market environment means identifying customers the utility wishes to keep, which customers it is willing to lose and which customers it would like to win over from its competitors. Furthermore, the utility must determine how best to keep those customers it wishes to keep, and how best to win over new customers. All of this implies a detailed knowledge of the customer base.
How can a utility develop this detailed knowledge of its customer base? The most cost-effective means is through the use of Geographic Information Systems (GIS). GIS allows utilities to ÒtagÓ any data set that has some kind of spatial reference, such as an address or postal code, to a specific customer. A great deal of useful information is available to utilities, from within their own organization and externally, including detailed demographic data, that can provide valuable clues to the services a customer might want from their utility. Developing this type of GIS system involves extracting and manipulating utility data, then adding readily accessible information from external agencies. This article provides a general description of the process and how the data sets are used. An example of a GIS system designed specifically for use in integrated resource planning was developed for Ottawa Hydro, and is discussed in detail below.
Using GIS for Marketing and Integrated Resource Planning
Existing utilities are in a unique position in the marketplace, because they have a wealth of useful market information locked away in their mainframes. They may choose to use this information for their own purposes in the event they wish to compete as an electricity service provider. Or, for municipal or other distribution utilities that do not generate their own power, they may wish to supply market information to brokers or act as brokers themselves, seeking to purchase electricity in bulk at the lowest cost for their customers.
Utilities must continue to provide electrical energy (and possibly other energy) reliably and at the lowest possible cost. This is essential to ensure credibility with both customers and investors and/or the public. Accomplishing this involves balancing supply and demand through integrated resource planning. Demand management is a legitimate means of avoiding potentially expensive new supply options and can enhance the utilityÕs image with its customers by being seen as efficient, cost-effective and ÒgreenÓ. Effective demand management necessarily implies a need for detailed knowledge about how customers use electricity, and how they will respond to demand management marketing initiatives. Utilities must be able to effectively assess the cost-effectiveness of demand management initiatives against supply alternatives and competitive initiatives from other utilities.
Developing the GIS Database
In order to keep supply costs low, and maximize revenue through maintaining or ÒcapturingÓ customers that will provide the utility with a high revenue generating potential, the utility needs to integrate information from a variety of different sources into its database:
1. Customer energy consumption and demand information from the utilityÕs own records.
2. Cost of supply information, including generation, transmission and distribution costs;
3. Customer demographic information, including number of occupants by household, their age, income, education levels achieved, and employment status;
4. Comparative energy use intensities (to assess which commercial, industrial or residential customers offer the greatest potential for efficiency improvements).
5. Other market profile data from market studies. These sources can provide additional segmentation of the market in order to assist in the development of marketing strategies.
6. Customer location, and total number of customers by geographic region.
With this information, utility planners can see where a customer is located, track the cost of supplying that customer through the transmission and distribution system (transformer station, substation, circuit), assess their revenue generating potential, likelihood to respond to different demand management and marketing initiatives, and determine total market potential within a specified geographic area.
Customer energy demand and consumption information can be downloaded from the utilityÕs own database. Other account information from the utilityÕs database is also downloaded in order to assist in customer identification. Specifically, the full service address and the customerÕs designated standard industrial classification (SIC) code. The SIC code is used to determine customer sector (residential, commercial, industrial) and sector (retail, single family detached home).
In order to make the information useful, standard billing data from the utilityÕs mainframe has to be downloaded and manipulated in such a way as to provide a breakdown of consumption and demand by calendar month for one full year. Billing data must usually be normalized to calendar months, as consumption data is not normally recorded at regular one month intervals for residential and small commercial customers. Consumption and demand information normalized to calendar months allows for an analysis of consumption by end-use. For example, heating and cooling load can be quickly identified by examining annual load profiles. Base consumption and demand is identified as occurring in the spring and fall, where there is assumed to be no or low heating and cooling energy consumption and demand. Consumption and demand above this level in the winter is assumed to be heating load, and consumption and demand above this level in the summer is assumed to be cooling load1. Baseload can be broken out into end-uses by applying a combination of standard end-use intensity (EUI) indices and expert analysis or special study data from the utility itself.
The net result is not a perfect representation of load by end-use for each customer, but a reasonable approximation. This information can be stored in tabular form and tagged to each customer account and location. This is accomplished by georeferencing customer account information through postal codes or address ranges. Customer location is quickly identified and can be viewed on a map, with utility transformer station boundaries displayed (see Figure 1). Transformer station boundaries can be added as a map layer into the GIS system, then used as a Òcookie cutterÓ to isolate accounts serviced by the station. The same approach can be used to isolate accounts for any geographic area desired. Circuit diagrams can also be added to improve the utilityÕs understanding of the cost to supply.
In essence, a utility can determine whether a particular customer or group of customers are placing significant demands on circuits, substations and transformers through high peak loading activity. If so, the marginal cost of supplying these customers can be easily estimated. Utilities can then measure customer load profiles against the marginal costs of electricity supplied. The net result would be a detailed, customer by customer display of revenues versus costs by customer. Utilities can then determine a number of things:
1. Demand management measures needed to balance load at the lowest cost;
2. Net revenue generated by each customer.
Current utility Automated Mapping /Facilities Management (AM /FM) systems are not designed to provide this type of analytical capability. However, information from an AM/FM system can be easily imported into a GIS system designed to provide market analysis information.
Demand Management Measures Needed to Balance Load at the Lowest Cost
If a utility is facing a situation where transmission/distribution system up-grades are required to meet current demand or projected increases in demand in the near term, it can enter into an integrated resource planning exercise. In this exercise, the cost of demand management measures are measured against the utilityÕs transmission/distribution system avoided-costs for meeting projected increases in demand. Demand management measures normally considered would be those that would reduce load during periods of peak demand, or shift it to off-peak periods. The choice of demand management measures will depend upon the nature of customer loads, the cost of implementing the demand management measure and its anticipated chances of success. It is with these last two points that GIS is a useful tool. GIS can greatly increase the penetration rate of demand management programs and therefore decrease its costs.
A GIS system was developed for Ottawa Hydro that was designed to assist in an integrated resource planning exercise for an overloaded transformer station. The utility had detailed daily load profiles for the transformer station, but was unable to identify specific customers serviced by the station, or to accurately and cost-effectively obtain detailed end-use data for the station.
A GIS system was used to isolate those accounts serviced by the transformer station and to provide detailed end-use profiles of each individual account. Once information had been broken down into end-use categories, the information could then be displayed graphically on a map (for example, the location and size of water heater or commercial cooling loads throughout the stationÕs service area).
The result was a detailed profile of the utilityÕs customers, including how, where and when load was being generated by specific end-uses. This information could then be used to target specific customers for load shifting programs.
Demographic information for the area was readily available. The transformer station services a wide variety of commercial, institutional and residential customers.
There were three very distinct residential areas within the service area, including low-income, middle to high income earners and one very wealthy neighbourhood. One of the most striking spatial representations of customer load made as part of the Ottawa Hydro study was of size of residential load by geographic region. The low, middle and high income areas were distinctly visible, just on the basis of electrical energy consumption (see Figure 2).
A result that came from this analysis was that the utility could target a very distinct area of middle-income customers in order to achieve the largest shifting of load to off-peak periods. This group represented customers with significant individual loads and a reasonable degree of cost-consciousness. These characteristics make them a cost-effective target group for a load shifting program.
ÒBEST CUSTOMERÓ IDENTIFICATION
As noted above, customers a utility or broker will want to attract are those that will generate the largest margins. These customers will not necessarily be large power consumers, but will instead be customers where revenues exceed the marginal cost to supply by a significant amount. Identifying these customers and determining how to keep or attract them will be the challenge for utilities and/or brokers. The important thing to do, once these customers have been identified, is to determine their energy service preferences. These preferences may include the provision of detailed energy consumption information, including breakdowns of that customerÕs consumption by end-use, developed as part of the GIS system to the customers themselves.
Residential customer segmentation, in which customers are categorized based upon socio-economic characteristics is a science in and of itself, and is beyond the scope of this article. However, research has been conducted in the U.S. in order to better understand consumer choices in a competitive utility industry environment. This research has segmented customers into specific groups with similar behaviour characteristics (for example, ÒbrandÓ loyalty) and preferences (those with a strong environmental ethic or those who are highly price or service conscious). Being able to identify customers with significant profit generating potential and knowing how to keep them loyal will provide utilities and brokers with a significant advantage over their competitors.
Demographic information is readily available for small geographic locales called enumeration areas. Enumeration areas vary in size, defined as the geographic area canvassed by one census representative and are the smallest geographic unit for which statistical information is widely available. Since people with similar socio-economic characteristics tend to live in the same neighbourhoods, enumeration area statistics can provide a great deal of useful information about customers.
Characteristics that are commonly used to characterize consumers include age, gender, income and education. Other factors are also important, including dwelling type, marital status and number of children. This information can then be linked to other information in a GIS system. This demographic information can be used to construct socio-economic profiles of customers, which can in turn be used to determine how they are likely to react to various utility programs, such as demand management initiatives, designed to meet their expectations and keep them loyal. The information can also be used to further refine their energy end-use consumption profile (for example, by profiling their plug loads). Commercial and industrial customers can also be characterized in a similar fashion. This type of research has been used with great success by marketing agencies selling any number of consumer products. Electricity services will soon be another commodity for sale in the market place, making this type of analysis extremely useful in a competitive environment.
Footnote: 1. Although there are other possible loads, such as pool pumps and heaters.
Richard Parfett is with R.S. Parfett & Associates located in Gloucester, Ontario. ET